RE: [LINUX-ABIT] It was Abit's experiment after all!

[William B. Clay]

Rod Roark wrote:

> And back to the BP6 and its problems with UDMA66 -- why are so 
> many folks hell-bent on using UDMA66?  The drives aren't that 
> fast anyway, it can't possibly make much difference.

You don't have to want use UDMA66 to use the third and fourth IDE 
channels on the BP6.  I just want three drives operating at 
UDMA33, each on its own IDE channel, for a software RAID5 array.  
I've tried 2.2.14, 2.2.15, and 2.3.51, with and without the 
latest IDE patches, and have yet to attain stable operation with 
heavy simultaneous load on both ide2 and ide3 (the HTP366 
channels).  If I run only one of the HPT channels, things are 
better but not perfect.  Problems: lost DMA interrupts, CRC 
errors, random seek errors.  The drives are brand new IBM Deskstar
34GXPs running UDMA33, and I have no problem with the one running 
on ide1 (Intel IDE controller).

--
=-          To unsubscribe, email [EMAIL PROTECTED] with the       -=
=-                body of "unsubscribe linux-abit".                 -=